HRV Analysis under the usage of different electrocardiography systems(Methodical recommendations)
These methodical recommendations are prepared according to the order of the Committee of Clinic Diagnostic Apparatus and the Committee of New Medical Techniques of Ministry of Health of Russia (protocol №4 from the 11-th of April,
2002) by the following authors: R.M.Bayevsky (Chair), G.G.Ivanov (Chair Deputy), L.V.Chireykin (Chair Deputy), A.P.Gavrilushkin, P.Ya.Dovgalevsky,
U.A.Kukushkin, T.F.Mironova, D.A.Priluzkiy, U.N.Semenov, V.F.Fedorov, A.N.Fleishmann, M.M.Medvedev (Secretary General)
Annotation
Current methodical recommendations on HRV analysis conclude perennial experience of the domestic research in this area. Presented materials are also con-sidering some foreign experience. These recommendations are relevant only for the so-called “short” records of heart rhythm (from few minutes to few hours) and should not be used for the 24-hour records.Here are presented generic working definitions and theoretical foundations HRV Analysis’ method. Also shown areas of method’s application and reasons for its usage. Standard methods of information gathering and recommendations on the following analysis are also proposed. Description of the basic methods HRV anal-ysis are also given and shown the ways of their standardization and further devel-opment.
Some basic methods of HRV Analysis’ results evaluation are also presented, including clinical-physiological interpretation and functional states evaluation. As-pects discussed of reproduction and comparability of received results. Also dis-cussed the prospects of following development of the methods of HRV Analysis.
Introduction
HRV Analysis has started to develop in USSR at the beginning of the 60s. One of the most stimulating reasons for its development was the success of space medicine research (ParinV.V., Baevsky R.M.,Gazenko O.G., 1965). In 1966 in Moscow the first symposium on HRV took place (on the mathematical analysis of heart rhythm) (ParinV.V., Baevsky R.M.,1968). Maximum activity of the research in HRV area in the USSR was noticed at the beginning of 70s- 80s (Zhemajtite D.I., 1965, 1970; Nidekker I.G., 1968; Vlasov Yu.A. and oth.,1971; Kudryavceva V.I., 1974; Voskresenskij A.D., Ventcel M.D., 1974; Nikulina G.A., 1974, Baevsky
R.M., 1972, 1976, 1979; Vorobev V.I., 1978; Kletskin S.Z., 1978; Bezrukih M.M., 1981; Gabinsky Ya.L., 1982).
The results of all those researches were gathered in 1984 monograph (Baevsky R.M., Kirillov O.I., Kletskin S.Z., 1984). The fast increase of the number of HRV researches was noticed during the last fifteen years in Western Europe and USA. For the last 5-6 years annually there were published up to few hundred works. In Russia after noticeable decrease of research activity in HRV Analysis at the end of 80s - beginning of 90s, the last few years also admitted growing attention to this method.
However in present time most of the Russian researchers use standards of measurements, physiological interpretations of HRV, and recommendations on the usage of this method, developed in 1996 by European Cardiology Society and North-American Electrophysiology Society, which are completely ignore enormous experience of domestic science.
Analysis of the considerable amount of publications in Russian magazines, materials of many conferences and symposiums show that studies of Russian scien-tists in HRV Analysis area are not only keeping pace with the western researchers, but in many points are being on the leading edge. Only for last few years in Russia have been published 4 monographs on HRV (Ryabikina G.V., Sobolev A.V., 1998, 2001; Mironova T.F., Mironov V.A., 1998, Flejshman A.F., 1999; Mihajlov V.M., 2000). Periodics regularly publish reviews on different aspects of HRV analysis (Ryabikina G.V., Sobolev A.V., 1996, YavelovI.S., Graciansky N.A., Zujkov Yu.A., 1997; Baevsky R.M., Ivanov G.G., 2001). Results of the studies of Russian scientists on HRV are regularly presented at the Russia-wide and International Car-diology congresses and symposiums (1996, 1997, 1999, 2002).
Current recommendations are developed on the basis of accumulation of ex-perience of domestic researches in this area with consideration of the data received abroad. These recommendations are not to be a literature review and completed with a very limited number of references, used in the text. Recommendations do not include materials on the clinical usage of the method. Their major goal is to stan-dardize methods of research and the ways to analyze data that is results of different studies could be comparable to each other.
Considerable amount of tools for HRV analysis have been developed and published by different firms and companies in Russia. Unfortunately, each of pro-ducers has tendency to use its own standards, based either on standards given in Eu-ropean-American recommendations, or developed by certain medical customers. This situation makes impossible to compare results of researches done with the us-age of different tools. Considering that there is a high probability of active and wide implementation of HRV analysis in Russia in the nearest future, we should think of the certain measures to standardize the method.
According to the decision of Commission of the diagnostic tools and equip-ment of the Committee on the new medicine technique of Russian Ministry of Health (protocol No 4 of 11.04.2000) there was created a group of experts to devel-
2
op methodical recommendations on the methods of HRV Analysis. The following recommendations are the one of results produced by this group and are relevant only to analysis of so-called “short” records of the heart rhythm using serial electro-cardilographical systems produced in Russia. Major items of this medical instruction realized in the following serial electrocardilographical systems produced in Russia:
1. Hardware-software complex “Varicard” (Institute of implementation of the new medical technologies, Ryazan);
2. Computer systems “Vita-Rhythm”, “VNS-Rhythm”, “VNS-Vita”, and “VNS-Spectrum” (firm “Neurosoft”, Ivanovo)
3. Computer electrocardiograph “Kardi” (firm “Medical Computer Sys-tems”, Zelenograd)
4. Hardware-software complex APK-RKG (“Mikor” company, Chelyabinsk)
5. Electrocardiographical complex “MKA01” and reograph tool “RP-KA2-01” with cardiograph channel (“Medass”, Moscow)
6. Complex of daily monitoring ECG “Cardiotechnika”(“INCART”, St Petersburg).
All above-mentioned hardware-software complexes suppose to work along with computer and provide generation of dynamical series of cardiointervals with the signal sampling rate up to 1000 Hz and higher. The accuracy of duration mea-surements is +-1 ms.
1. Basic working definitions.
HRV Analysis is a method of evaluation of the state of the mechanisms of the regulation of physiological functions in human organism and animals, partially of the common activity of regulatory mechanisms of the neurohumor regulation of heart correlations between sympathetic and parasympathetic departments of vegetative nerve system.
Current activity between sympathetic and parasympathetic departments is a result of the reaction of multi-contour and multi-circle blood system regulation, changing its parameters for achievement of optimal adaptive response, which re-flects adaptive reaction of whole organism.
Adaptive reactions are individual and can be realized for different persons with different level of participation of functional systems, which possess in their turn the backward connectivity, changing in time and having changeable functional orga-nization. The method is based on the recognition and measurement time intervals between R-picks ECG (R-R intervals), building dynamic series of cardiointervals and following analysis omitted numerous lines with the different mathematical meth-ods. Dynamic series of cardiointervals are called as cardiointervalogramm (CIG).
A dynamical series can be considered as stationary and nonstationary. Ran-dom processes are called stationary if they proceed approximately homogeneously and look like continuous oscillations about a mean position. Stationary processes
3
are characterized by ergodicity, i.e., averaging over time corresponds to averaging over an ensemble. In other words, we should obtain identical parameters at any time interval. Virtually each time interval contains some elements of nonstationarity (fractal components). In these methodical recommendations cardiointervalogramm is considered as a steady-state random process with the relevant interpretation of the data resulted from its analysis. In the recent years, the methods of nonlinear dynam-ics have been extensively applied to evaluate the fractal components of time inter-vals (Goldberger A., 1991; Fleishmann A. N., 1999, 2001; Gavrilushkin A.P., Masluk A.P., 2001).
During the analysis of the dynamic series of cardiointervals we should distin-guish between short-time (“short”) and long –time (“long”) records. Under the last, as a rule, meant data received from 24-hour and 48-hour monitoring ECG (Holters monitoring). Under so-called “short” records meant data received during the re-searches lasting minutes, tens of minutes, or hours.
Dynamic series of cardiointervals can be obtained from analysis of any of cardiographical records (electrical, mechanical, ultrasound, etc), however in the cur-rent document we are focused only at the data resulting analysis of electrocardiosig-nals.
VHR Analysis includes three stages:1. Measurement of the length of R-R intervals and presentation of the dy-
namic series of cardiointervals in a form of cardiointervalogramm.2. Analysis of the dynamic series of cardiointervals.3. Evaluation of the results of VHR analysis.Measurement of the length of R-R intervals is executed by hardware or soft-
ware methods with the accuracy up to 1ms. The problem of recognition R-picks of ECG in different hardware-software complexes is solved differently. The presenta-tion of dynamic series of cardiointervals can be realized in the numeric or graphical view.
The methods of analysis of dynamic series of cardiointervals can be divided
into visual and mathematical. Visual analysis of cardiointervalogramms (rhythmo-
gramm) has been developed by D.Zhemajtite (1965, 1972). The classification of
rhythmogramms proposed by her has being actual up-to-date (T.V.Mironova,
V.A.Mironov, 1999). The mathematical methods of the analysis can be divided into
three large classes:
• general variability research (statistical methods or analysis in time domain).
• research of HRV periodic components (frequency analysis).
4
• research of internal organisation of a dynamic sequence of cardiointervals
(autocorrelated analysis, correlation rhythmography, methods of non-linear dynam-
ics).
Numerous values (parameters of HRV) obtained as a result of HRV analysis
are estimated in different ways by various explorers depending on the scientific-the-
oretical concept used.
5
мс
с
Pic.1. Formation of cardiointervalogram (CIG) while inputing electrocardi-
ography signal. Upper curve represents electrocardiogram (ECG), bottom curve
shows CIG (ordinate axis represents duration of cardio intervals in milliseconds;
abscissa axis represents check-in time of cardio intervals (hours, min, sec.).
Points mark elements of CIG corresponding to intervals between RR-fingers of
the ECG.
2. SCIENTIFIC AND-THEORETICAL BASIS OF A METHOD
The main information about systems regulating a heart rhythm is included in
«dispersion functions» of cardio intervals durations. It is necessary to take into ac-
count current level of blood circulation function system. Analysing HRV the ques-
tion is a so-called sinus arrhythmia, which represents complex processes of interac-
tion of different circuits of cardiac rhythm regulation. If a rhythm disturbance of dif-
ferent origin appears the application of special methods on recovery of stationarity
of studied process is required or it is necessary to use the special analytical ap-
proaches.
A dynamic sequence of cardio intervals can be analysed and estimated on the
basis of different scientific and theoretical conceptions. Depending on scientific or
practical tasks it should be recommended to use one of the following three ap-
proaches:
1. To consider the changes of a cardiac rhythm in connection with adaptive re-
action of an organism as a whole, i.e. like a development of different stages of a
general adaptation syndrome (G.Selie, 1961).
2. To consider the oscillations of cardio intervals durations as an outcome of
influence of multi-circuit, hierarchically organized multi-level managing system of
physiological functions of organism. This approach is based on principals of biolo-
gical cybernetics (V.V. Parin, R.M. Bayevski, 1966) and on theory of functional
systems (P.K. Аnokhin, 1975). In that case it is possible to consider changes of
6
heart rate variability parameters to be specified by formation of different functional
systems which coincide to outcome required at the moment.
3. To consider the changes of cardiac rhythm in connection with activity of
neurohormonal regulation mechanisms as a result of activity of different links of ve-
getative nervous system.
Nowadays the theory of adaptation is one of the fundamental directions of
modern biology and physiology. The adaptive activity of human and animal organ-
ism provides not only survival and evolutional development but daily adaptation to
changes of environment as well.
G.Selie's theory on common adaptation syndrome describes phase nature of
adaptive reactions and founds a leading role of consumption regulatory systems un-
der acute and chronic stress in development of majority of pathological conditions
and diseases. Blood circulation can be considered as the sensing indicator of adapt-
ive reactions of organism as a whole (V.V.Parin and co-authors, 1967). Variability
of cardiac rhythm represents well degree of regulatory systems strain corresponded
by both activation of pituitary gland – adrenal system and reaction of sympathoad-
renal system arising in reply to any stress effects.
Detailed study of HRV by using methods of autocorrelated and spectral ana-
lysis has resulted in developing the approach based on postulates of biological cy-
bernetics and theory of functional systems. The concept of heart rhythm variability
as a result of influence of numerous regulatory mechanisms (nervous, hormonal,
humoral) on blood circulation is in the basis of this approach.
Functional system of regulation of blood circulation represents multiple-circuit,
hierarchically organized system in which the predominant role of separate links is
determined by current needs of organism. The most simple double-circuit model of
regulation of cardiac rhythm is based on the cybernetic approach which suppose that
system of sine node regulation can be represented by two interdependent levels (cir-
cuits) namely central and autonomous with direct coupling and feedback loop (see
pic.2). In that case the effect of autonomous level (circuit) is identified with respirat-
7
ory arrhythmia and the effect of central circuit is connected with nonrespiratory ar-
rhythmia.
Humoral channel of regulation Cortex
High vegetative centres and subcortical nervous centres Cardiovascular centre of oblong brain Cores of vagues nerves
Respiratory tract Sine node
Lungs
CENRAL CIRCUIT AUTONOMIC CIRCUIT Nervous channel regulation
Pic. 2. Scheme of double-circuit model for cardiac rhythm regulation.
The working structures of autonomous circuit of regulation are: sine node
(SN), vague nerves and their cores in medulla (circuit of parasympathetic regula-
tion). Thereby breathing system is considered as a member of a feedback loop in au-
tonomous circuit of cardiac rhythm (СR) regulation.
The activity of central regulation circuit, which is identified with sympathoad-
renal influences on a heart rhythm, is connected to non-respiratory sinus arrhythmia
(SA) and is characterized by different slow-wave components of cardiac rhythm.
The direct coupling between central and autonomous circuits is fulfilled due to
nervous (basically sympathetic) and humoral communications. The feedback is
provided by afferent impulsation from baroreceptors of heart and vessels, from
chemoreceptors and vast receptor zones of different organs and tissues.
Autonomous regulation in the state of rest is characterized by availability of
expressed respiratory arrhythmia. Respiratory waves strengthen during the dream
when the central influences on autonomous circuit of regulation decrease. Different
loadings on organism, which require engaging of central circuit of regulation in CR
control procedure, lead to the attenuation of respiratory component of SA and to the
intensification of its non-respiratory component.
8
Central circuit of CR regulation is a complex multi-level system of neuro-
humoral regulation of physiological functions, which includes numerous links from
subcortical centres of oblong brain up to hypothalamus-pituitary level of vegetative
regulation and brain cortex. Structure of this system of regulation can be schematic-
ally represented by consisting of three levels. Certain functional systems or levels of
regulation correspond to these levels rather then anatomic-morphological structures
of brain. Those systems or levels are as follows:
1-st level provides organization of co-operation of organism with an environ-
ment (adaptation of organism to exposures). The central nervous system, including
cortical mechanisms of regulation, which synchronize functional activity of all sys-
tems of organism according to effect of environmental factors, fall into this level
(level А).
2-d level executes equilibrium between different systems of organism and
provides an intersystem homeostasis. The basic role in this level is played by
highest vegetative centres (including hypothalamic-pituitary system), which
provides hormone-vegetative homeostasis (level B).
3-rd level provides intrasystem homeostasis in different systems of organism,
in particular in cardio-respiratory system (the system of blood circulation and sys-
tem of breathing can be considered as one functional system). Here the leading role
is played by subcortical nervous centres, in particular by vasculomotor centre as a
part of subcortical cardiovascular centre stimulating or depressing heart by means of
fibres of sympathetic nerves (level C).
Non-respiratory SA represents oscillations of СR with the periods bigger than
6-7 seconds (below 0,15 Hz). Slow (non-respiratory) oscillations of cardiac rhythm
correlate with similar waves of arterial pressure (AP) and of plethysmogram. Slow
waves of 1-st, 2-nd and higher orders are distinguished by observers. СR structure
includes not only vibrational components in the form of respiratory and non-respirat-
ory waves but also non-periodic processes (so-called fractal components).
9
Origin of these CR components is being connected with multi-level and non-
linear nature of processes of cardiac rhythm regulation and with availability of tran-
sient processes. The heart rhythm is not strongly stationary casual process with er-
godic properties, what entail repetition of its statistical characteristics on any arbit-
rary taken stretches.
Heart rate variability reproduces complex picture of diverse control influences
on system of blood circulation with interference of periodic components of different
frequency and amplitude and also with non-linear co-operation of various control
levels.
By using CR records with duration less than 5 minutes we artificially limit the
number of regulatory mechanisms (control circuits) studied, we narrow down range
of studied control influences. The longer is the sequence of cardio intervals being
analysed, the bigger the number of regulatory mechanism levels could be investig-
ated.
Approach to the HRV analysis is based on the idea of concerning mechanisms
of neurohormonal regulation and is the closest and clearest to the physiologists and
especially clinicians. It is known, that heart rhythm is regulated by vegetative, by
central nervous system and by several humoral and reflex effects. Parasympathetic
and sympathetic nervous systems are in certain co-operation and both are affected
by central nervous system and several humoral and reflex factors.
There is a permanent effect of sympathetic and parasympathetic influences at
all levels of regulation. The real relations between two departments of vegetative
nervous system are complex. As a matter of fact is that there is a different degree of
activity of one of the departments of a vegetative system while changing activity of
another. It means, that actual heart rhythm can sometimes be the simple sum of sym-
pathetic and parasympathetic stimulation, but sometimes sympathetic or parasym-
pathetic stimulation can interact in a complicated way with origin parasympathetic
or sympathetic activity.
10
The decrease of activity in one part of autonomic nervous system and in-
crease it in another is observed at achievement useful adaptative result.
For example, excitement of baro-receptors with increase of arterial pressure
results to decrease of frequency and force of intimate reductions. This effect is
caused by increase of parasympathetic and reduction of sympathetic activity. Such
type of interaction corresponds to a principle "functional synergy". It is necessary to
emphasize, that the described various approaches to the analysis HRV are comple-
mentary. The current activity of sympathetic and parasympathetic parts of autonom-
ic nervous system is system reaction of multiplanimetric and multilevel system of
regulation.
3. Basic areas of application of the method and indication to its use
During 40-year's term of application of various methods of HRV in the differ-
ent areas of physiology and clinical medicine, the sphere of their use continues to
extend with each year. It is essentially important that the analysis of HRV is not
specialized method of diagnostic. It is possible to list some examples, where it is ap-
plied to specification of the diagnosis of the certain diseases. In particular, it is diag-
nostics of neuropathy at diabetes mellitus. In many cases it is study of nonspecific
reactions of organism on influence of the various factors or at the certain diseases.
Preceding from the submitted scientific - theoretical rules it is possible to allocate
conditionally four directions of application of methods of the analysis HRV:
1. Estimation of functional condition of the organism and its changes on the ba-
sis of definition of parameters autonomic balance and neurohumoral regula-
tion;
2. Estimation express of the adapted answer of organism at influence various
stresses;
3. Estimation of a condition of separate parts vegetative regulation of circulation
of blood;
11
4. Development prognosis on the basis of an estimation of the current functional
condition of organism, express it the adapted answers and condition of sepa-
rate parts regulation of the mechanism.
The practical realization of the specified directions opens a boundless field
of activity, both for the scientist, and for physicians. The rough and rather incom-
plete list of areas of use of methods of the analysis HRV and indications to their
application made on the basis of the analysis of the modern domestic and foreign
publications is offered below.
1. Study of autonomic regulation of heart rhythm at the practically healthy
people (initial level vegetative regulation, vegetative reaction, vegetative
providing with activity);
2. Estimation vegetative regulation of a rhythm of heart at the patients with
various diseases (change vegetative balance, degree of prevalence of one
of the departments vegetative of nervous system). Reception of the addi-
tional information for diagnostics of some forms of diseases, for example,
independent neuropathy at diabetes mellitus.
3. Estimation of functional condition of the regulation systems of organism
on the basis of the integrated approach to system of circulation of blood as
to the indicator adaptation of activity all organism.
4. Definition of vegetative regulation (vago-, norma- or sympathotonia).
5. Prognosis of risk of sudden death and fatal arrhythmias at a heart attack
myocardium and IDH, at the patients with gastric infringements of a
rhythm, at chronic intimate insufficiency caused arterial hypertension, car-
diomiopathy.
6. Allocation of groups of risk on development of menacing life of the raised
stability of an intimate rhythm.
7. Use as a control method at realization of various functional tests.
8. Estimation of efficiency of medical-preventive and improving measures.
12
9. Estimation of a level of stress, degree pressure of regulation systems at ex-
treme and subextreme influences on organism.
10. Estimation of functional condition of the man - operator.
11.Using for estimate of functional condition at mass preventive inspections
of different quota of the population.
12.Prognosis of a functional condition (stability organism) at professional se-
lection and definition professional of suitability.
13.Monitoring HRV in surgery with the purpose of the express objectivity of
operational stress and control of adequacy anaesthesia, and also for a
choice such as and dosages of anesthetic protection and for the control in
after surgical operation period.
14.Estimation of reactions autonomic nervous system at influence on organ-
ism of electromagnetic fields, intoxications and others pathogenic factors;
15. Choice of optimum therapy in view of a background of autonom-
ic heart regulation, in the control of efficiency of treatment, correction of doze
of preparation;
16. Estimation and prognosis of mental reactions by expressed vege-
tative background;
17. Using this method in neurology for an estimation of a condition
autonomic nervous system at various diseases;
18. Control of a functional condition of the organism in sports;
19. Estimation of the vegetative regulation during development of
children and teenagers. Application as a control method in school medicine
for social - pedagogical and medical psychological researches;
20. Control of the functional condition of a foetus in obstetrics. Ap-
plication in the neonatal period of development of the organism.
The above-mentioned list is not exhaustive. It will constantly increase. The
basic affidavit to the application of methods of HRV analysis is probable changes of
organism’s system regulation, in particular changes of vegetative balance. As there
13
are practically no such functional conditions or diseases, in which the mechanisms
of vegetative regulations would take part, then the sphere of application of HRV
analysis is really inexhaustible. It means, that nowadays this method is the only
available, non-invasive, simple enough and relatively cheap method for estimation
of vegetative regulation. Taking into consideration the vast perspectives of the de-
velopment of the method, it is important to provide its standards and comprehension
of data obtained by different researches.
4. Main medical and technical conditions.
4.1. Conditions for continuity of registration of heart rhythm.
The continuity of CR registration depends on purposes of research. The dura-
tion of records can be at range from several minutes to several hours. For instance,
at mass prophylactic examination or at preliminary policlinic and clinic researches
5-minutes registration of ECG is applied. At functional tests the continuity of regis-
tration can be at range from 10-15 minutes to 1.5-2 hours. Control tests from 3 to 5
hours can be in need during the surgical operations, and at last the continuity of per-
sistent registration can reach 10-12 hours in resuscitation departments and at re-
search of a dream. It is supposed to be pointed four kinds of HRV investigation:
1. Transitory (operative or survey) records (standard continuity – 5 minutes)
2. Records with middle continuity (up to 1-2 hours)
3. Long records (up to 8-10 hours)
4. Daily records (24 hours and more continuing)
Definite tasks can require shorter periods of time records (1-2 minutes). In
these guidelines long and daily records are not studied. As for the records of middle
continuity, in this case their application is supposed in the frames of holding the
functional tests.
In spite of continuity of registration during the data analysis as basic selec-
tion, it is recommended to use 5-minutes segments of record. In separate cases it is
allowed to use shorter selection by work with highly stationary processes (emotional
stress, steady phase of physical training). It is worth to use in each stationary stage
14
5-minute standard segments of record and sum up in a proper way the results of
analysis of these segments, if the necessity of estimation of cardiointervalogram by
continuing observation will arise. The analysis of more prolonged segments of
record require special elaboration, because while estimating them one must take into
account presence of periodical components in their contain, reflecting the state of
higher level regulation and it is also important to pay special attention to the stability
of functional condition and transitional processes.
4.2.The methods of HRV investigation.
The investigation of HRV can be parallel or specialized. In first case it is held
simultaneously with ECG registration, EHO-CG for the purposes of diagnosis and
medical control or during Holter monitoring. In the second case it is purposeful in-
vestigation of HRV with utilization of specialized systems.
According to this, four kinds of investigations can be pointed:
a) operative investigations in condition of relative rest;
b) investigations during functional tests;
c) investigations in condition of usual activity;
d) investigations in clinical conditions.
Each of these kinds of investigations is characterized by peculiarities of meth-
ods.
4.2.1. Operative investigations in condition of relative rest.
ECG signal is registered in one of the standard (better in the 2-d or 3-d) chest
sections. The duration of record must be not less than 5 minutes. It is better to hold
record not less than 10 minutes, if there are disorders of rhythm. Analysis of the 2-d,
3-d consistent 5-minutes’ records confirms the state of physiological status stability.
HF must be known in experimental and clinical investigations for proper comparison
of receiving data.
HRV investigation must be held not earlier than in 1.5-2 hours after meal in a
quiet place, constant temperature of which is 20-22°. It is necessary to revoke phys-
ical and therapeutic procedures and treatment with medicine before investigation.
15
Before the beginning the investigation, 5-10 minutes period of adaptation will be
necessary.
The record of ECG is being made in lying position, with calm breath. The
surrounding during the investigation must be calm. It is desirable to hold the investi-
gation with women during the mensal period, as hormonal changes in organism re-
flect themselves at cardiointervalogram. It is necessary to dismiss all obstacles lead-
ing to emotional excitement, not to speak with investigated and other, expel tele-
phone calls and appearance of other persons including medical workers. During
HRV investigation Patient must breathe without taking deep inhale, coughing and
swallowing saliva.
4.2.2. Investigations at functional tests.
Functional test is an important part of HRV investigation. The main aim is to
estimate of functional supplies of mechanisms of vegetative regulation. Different
links of system of management of physiological functions can be tested depending
on kind of functional work.
Sensibility and reactivity of vegetative nervous system, its sympathetic and
parasympathetic sections under the influence of one or another testing factor can
serve as diagnostic and prognostic criteria.
For example, in the case of diabetic neuropathy the reaction of parasympa-
thetic link of regulation on the trial with fixed tempo of breathing (6 breathes per
minute) is one of the most important diagnostic signs. The schedule of functional
tests, most often used in researches of HRV, is presented below:
a. Active and passive orthostatic test (when it is needed clinoorthostatic
test).
b. Test with the fixed breathing tempo.
c. Valsaw’s test.
d. Tests with maximal breathing delay on inhale and exhale.
e. Isometric cargo test.
f. Cargo tests at veloergometre.
16
g. Pharmacological tests (with β-blocators, atropine and other drugs).
h. Ashner’s test.
i. Sinocarotid test.
j. Psychophisiological test.
Represented schedule of functional tests is not full. Each of pointed tests is
led by its own special method. Depending on the type of test used the length of
recording CR can hesitate from some minutes (during the test of fixed breathing
tempo) to some hours (during pharmacological tests).
It is necessary to point out next specialities of HRV analysis during functional
tests:
• Background (initial) record must be led in condition of the rest (see above)
during not less than 5 minutes. For comparison with background record analogous
by length records, received on different stages of function test must be used.
• Transitional process during functional tests must be analyzed by special
methods (these methods are not looked through here). In this case it must be pointed
out from the record visually or automatically with the use of appropriating algo-
rithm, registrating instationarity and non-linearity of the process. The analysis of
transitional processes can have independent diagnostic and prognostic sense. Tran-
sitional process, depending on the type of functional tests can take shorter or longer
time.
• The estimation of changes of indices HRV by functional tests must be done
with the registration of data received by other methods of research.
4.2.3. The researches in conditions of ordinary activity or in the case of
making professional work.
The using of analysis HRV as the method of estimation of adapting possibili-
ties of the organism or current stress stage occur practical interest for different do-
mains of applied physiology, professional and sport medicine and for socio-ecologi-
cal researches. The development of donosologic diagnosis made it possible to point
out almost wealthy people, enormous group of people with high and very high ten-
17
sion of regulatory systems, with heightened risk of disruption of adaptation and ap-
pearing of pathologic deviation and diseases. Such people need regular control of
stress stage and recommendations for wealthy preservation.
The problem of chronic stress when there is constant elevating tension of reg-
ulating systems concern almost all the population, but particularly important for sep-
arate professional groups, whose work is united with the influence of complex of
stress factors. They are, in particular, operators of computer systems, controllers,
drivers and also businessmen and administrator-commanding apparatus. The analy-
sis of HRV is an adequate method of the estimation of the stress stage of their ev-
eryday work. Here depending on target the using of every of three types of research-
es (short-time, middle-time or long-time) is possible.
Short-time or operative researches with the recording lengths of 5-15 minutes
may take place in the system of common investigations when it is necessary to esti-
mate the group’s of people state and to point out persons with elevated risk of
pathologic process. In such researches parallel gathering of anamnesis, registration
of complaints, manner of life and anthropometric dates are important. The recording
must take place in the conditions of relative calm in laying or sitting position.
Middle-time records (till one hour) are expediency to carry out using to sepa-
rate stages of work. For instance, at the beginning and at the end of working day,
during the lesson, during making concrete working operation. In sport medicine
such records can take place before and after competitions, during making individual
physical trainings (only stationary sections of record). In the case of operator work,
it is controlled before and after relay.
Long time records are researches during working relay, during working day
and during night sleep.
The analysis of HRV in middle and long-time records is recommended to
take place using 5-minutes segments for the learning of the dynamics of adaptation
process. The checking of every analyzing segment on stationarity has essential
sense. The sections of record reflecting transitional processes must be analyzed with
18
the use of special methods. When the results of HRV analysis are estimated, condi-
tions of records, reaction factors and the position of investigated person (laying, sit-
ting, moving, etc.) must be registrated.
4.2.4. Researches in clinical conditions.
In application to clinical conditions it is also important to differ above-men-
tioned kinds of researches. Short-time researches must be inspected as operative, re-
viewing and previous. They can take place at the beginning and at the end of treat-
ment or regularly at the process of treatment for the definition of functional patient’s
state. The middle-time records, which are led because of functional tests, are the
most adequate to clinical conditions.
Moreover, such records take place in function with control of treating proce-
dures, for example, in the case of physiotherapeutic reaction. Researches from
surgery and anesthesiologic fields also concern to middle-time records. They are as
records made directly during surgery operations for the control of adequacy of anes-
thesia, as the control of the patient’s state at the nearest after-operation period.
Long-time operations are used for the HRV analysis at the after-operation pe-
riod and in resuscitation practice. The estimate of the stress level and opportune ex-
position of overloading and exhaustion of regulatory mechanisms plays the most im-
portant role for prevention of menace states and death. The researches of sleep led
in neurology and psychology are also the example of long-time records.
It is important to underline, that the specialty of HRV analysis in the case of
using this method in clinical practice is, that the physicians have to understand clear-
ly nonspecialty of results received and not to try to look for of HRV pathognomic to
one or another form of pathology. The dates of HRV analysis have to be compared
with other clinical data, instrumental, biochemical, anamnesis.
4.3. The requirements for software, standards of processing.
1. The presentation of previous data in a way of cardiointervalogram with possi-
bility of editoring (remove of arthefacts and extrasystols) has to b provided;
19
2. It is recommended to provide automatic discretion of arythmics and their in-
terpolation without breach of stationarity of dynamic row of cardiointervals;
3. It is recommended to provide the possibility of reform of dynamic row of car-
diointervalsto equidistant row with the frequency of quoting 250,500…..or
1000 (possibility to choose the frequency);
4. Possibility of choosing the analysis method;
5. Representing the results of analysis in a graphic way (varying pulsograms,
scaterograms, spectrums, etc.);
6. Forming of the table of results of analysis and appropriate graphic represent-
ing by all methods chosen of analysis;
7. It is recommended to provide the possibility of automatic estimate of analy-
sis’ results (mainly for commercial programs and mass consume);
8. Database for keeping of initial information (desirable also initial ECG-signal)
and results of analysis;
9. The possibility of receiving information (after the user’s inquiry) concerned of
structure of program, rules of working with it and interpretation of calculating
exponents must be provided;
10.Complementary requires can include the possibility of:
a) estimate of stationarity of dynamic row and rejecting of instationary sections;
b) successive analysis of selections of proposed capacity with proposed pace
(continuously sliding method);
c) the identification of cogs P, Q, S, T and segments PQ, ORS, QT and ST in
ECG and also lining up the dynamic row of meanings by set exponents.
5. Essential methods of HRV analysis.
5.1. Statistical methods.
These methods are used for direct amounting of estimation of HRV at investi-
gated period of time. During their use the cardiointervalogram is inspected as the
collection of successive time intervals, namely intervals RR. Statistic characteristic
of dynamic row of cardiointervals include: SDNN, RMSSD, PNN50, CV.
20
SDNN or CKO is a summarized index of variability of quantities of RR inter-
vals of all investigated period (NN means a row of normal intervals “normal to nor-
mal” except extrasystols);
CKO is neuter quadrate deviation (conveys in ms);
SDNN is standard deviation of NN intervals (CKO analog);
SDANN is standard deviation of neuter meanings SDNN from 5-minutes seg-
ments for middle-time records, long-time records or 24 hour records. Also standard
deviations of neuter meanings of other exponents can be reviewed by the same way;
RMSSD is quadrate root from the sum of quadrates from the difference of
quantities of successive pairs of NN intervals (normal RR intervals);
NN50 is number of pairs of successive NN intervals, which differ for more
than 50 milliseconds, received for all the record period;
PNN50 (%) is percent NN50 from the common number of successive pairs of
intervals which differ for more than 50 milliseconds received for all the record peri-
od;
CV is coefficient of variation. It is favorable for the practical use it constitutes
an estimation CKO;
CV=CKO/M*100, where M is a neuter meaning of RR intervals;
D, As, Ex are second, third and fourth statistic moments. D is CKO in
quadrate, it reflects summary power of all periodic and non-periodic fluctuations. As
is a coefficient of asymmetry, it let us judge about the stationarity of investigated
dynamic row, presence and expression of transitional processes including trends. Ex
is a coefficient of excessiveness, reflects the rapidity (precipice) of change of casual
non-stationed components of dynamic row and presence of local non-stationarity.
5.2. Geometric methods (variable pulsometry).
The essence of varying pulsometry confines in learning the allocation law of
cardiointervals as casual quantities. A variable curved line (curved line of allocation
of cardiointervals is histogram), is lining up in this case and its main characteristics
are definite by: Mo (Mode), Amo (Model’s amplitude), MxDMn (variable range).
21
Mode is a most often met meaning of cardiointerval in this dynamic row. Mo differs
a little from mathematic expectation (M) in the case of normal allocation and high
stationarity. Amo (Model’s amplitude) is the number of cardiointervals appropriated
to the quantity of Mode in percent to the capacity of selection. Variable range
(MxDMn) reflects the grade of variability of cardiointervals’ meanings in re-
searched dynamic row. It is calculated by the difference of maximal (Mx) and mini-
mal (Mn) meanings of cardiointervals and that’s why it can be distorted in the case
of arythiyes and arthefacts.
During lining up histograms (or variable pulsograms) the choosing method of
gathering data has the major meaning. A traditional manner of grouping cardiointer-
vals is the range from 400 to 1300 ms with the intervals of 50 ms was constituted in
perennial practice. Thus, 20 fixed ranges of cardiointervals’ length are pointed
which let us to compare variable pulsograms received by different researchers at dif-
ferent groups of researches. In this case the selection capacity, in which the group-
ing and lining the variable pulsograms up, is also 5-minute standard. Another way of
lining up the variable pulsograms confines in definition of modal meaning of car-
diointerval for the first and then using ranges for 50 ms to form histogram to both
sides from the mode. Variable pulsogram can be also presented by “plane” graphic
of density of allocation (look pic.3).
According to the data of variable pulsometry widely diffused in Russia index
of tension of regulatory systems or stress index is calculated.
Ин=AMo/2Mo*MxDMn
West European and American researchers use approximation of curved line
of cardiointervals’ allocation as a triangle and count a so-called triangular index –
the integral of density of allocation (common number of cardiointervals), referred to
the maximum of allocation density (AMo). This exponent is designated as TINN
(triangular interpolation of NN intervals).
22
Moreover, the lining up of histograms is used by differential meanings of ad-
joining cardiointervals with the approximation of their exponential curved line and
with the counting of logarithmic coefficient and other way of approximation
23
Pic.3. Models of variative pulsograms by tachycardia and normocardia (his-togram and diagram of allocation intensity)
5.3. Autocorrelating analysis.
Discount and lining up of autocorrelating function of cardiointervals’ dynamic
row is directed at the studying of internal structure of their row as of casual process.
Autocorrelating function is a graphic of dynamics of correlation coefficients re-
ceived at the case of successive deviation of analyzing dynamic row for one number
relatively to its own row.
After the first dislocation for one value the correlation coefficient is such less
than one as more breathing waves are expressed (pic.4). If slowly waved compo-
nents dominate in investigated selection, then correlation coefficient will be insignif-
icantly lower than one after the first dislocation. Next dislocations lead to gradual
extenuation of correlation coefficients. Autocorrelogram let us try on hidden period-
icity of CR.
C1 is the meaning of correlation coefficient after the first dislocation and C0
is number dislocation, which result to the meaning of correlation coefficient becom-
ing negative, are recommended to be used as amounting meanings of autocorrelo-
gram.
24
Pic.4. Models of autocorrelograms with expressed breathable waves (at the top), with the prevalence of slow (in the middle) and the slowest waves (at the bottom)
25
5.4. Correlating rhythmography-scatterography.
The essence of correlating rhythmography method is in graphic rendering of
successive pairs of cardiointervals (last and next) in by-measured coordinate plain.
During this quantity R-Rn is put on the axle of abscissa, and quantity R-Rn+1 is on
the axle of ordinates. The graphic and region of points received by such way
(Puankare’s or Lorence’s spots) is called correlating rhythmogram or scatterogram.
This way of estimation of HRV concerns to the methods of nonlinear analysis and it
is mostly useful for the cases when rare an sudden breaches (ectopical systoles and/
or “prolapse” of separate heart systoles) are met at the background of monotone of
the rhythm.
When lining up scatterogram a join of points whose centre is situated at the
bisector is organizing. The distance from the centre to the beginning of coordinate
axles suit to the linger of heart cycle (Mo) most waiting for. The value of deviation
of the point from bisector to the left shows, how is this heart circle shorter than the
previous one, to the right from bisector, how it is longer than the previous one. The
following indicators of scatterogrm are supposed to be calculated:
a. Length of the main “eddy” (without extrasystole and artifacts) corre-
sponds with variation range. With its physiological sense this indicator
does not differ from SDNN, i.e. it reflects the summary effect of HRV
regulation, but indicates maximum amplitude of fluctuations of R-R in-
tervals duration;
b. Width of scatterogram (perpendicular to the long axis drawn across its
middle –w);
c. Scatterogram area is to be calculated according to the ellipse area for-
mula:
S=(π·L·w)/4.
Normal shape of scatterogram is represented by ellipse starched
along bisector. This very arrangement of ellipse means that some non-res-
piratory arrhythmia value is added to the respiratory one. The round
26
shape of scatterogram means the absence of arrhythmia non-respiratory
components. Narrow oval (pic.5) corresponds to the predominance of
non-respiratory components in general rhythm variability that is deter-
mined by the “eddy” length (scatterogram length).
The oval length correlated well with HF magnitude and correlated
with LF (see below). In case of arrhythmias, when statistic and spectral
analysis methods for heart rate variability estimation are not very informa-
tive or not acceptable, it would be necessary to use correlation rhythmo-
gram estimation.
27
Pic.5. Models of correlative rhythmogram (CRG). Normal CRG (at the top), with a patient with arrhythmia (at the bottom).
5.5. Spectral methods of HRV analysis.
Spectral methods of HRV analysis are widespread nowadays. Analysis of
spectral density of oscillations rate gives information on power allocation depending
on the oscillations frequency. Spectral analysis application allows making quantita-
tive evaluation of different frequency constituents of the cardiac rhythm oscillations
and represents graphic correlations of different CR components that reflect activity
of definite regulatory mechanism links.
There are parametric and non-parametric of spectral analysis. Former in-
cludes autoregressional analysis, latter includes Fure’s quick transformation (FQT)
and periodogram analysis. Both groups of methods provide comparative results.
Parametric and, in particular, autoregressional methods requires correspondence of
an analyzed object with the definite models. General for all classical methods of
spectral analysis is the problem of Windowing application. Its main role is reducing
shift value in periodogram spectral evaluations. There are definite differences in
spectral estimation of data using periodogram method with even window (256 val-
28
ues of RR) and application of different inter segment shift of levels and of different
number of indication per segment.
Enhance of clearance in case of inter segment shift increase indication per
segment causes appearance of many additional peaks in the spectrum and increase
of peaks amplitude in the right half of spectrum. In application of spectral analysis
of HRV volume of analyzed selection has important meaning. In brief records (5
minutes) there are three main components outlined. These components correspond
to respiratory waves range and slow waves of 1-st and 2-d degree (pic.6).
In western literature the corresponding spectral components were called
“High Frequency” (HF), “Low Frequency” (LF) and “Very Low
Frequency” (VLF).
Frequency ranges of each of three above-mentioned spectral components are
discussional. According to Euro-American recommendations (1996) the following
frequency ranges are offered:
HFR (respiratory waves) – 0.4-0.15 Htz (2.5-6.5);
LF range (slow waves of the 1-st degree) – 0.15-0.04 Htz (6.5-25);
VLF range (slow waves of the 2-d degree) – 0.04-0.003 Htz (25-333).
In case of long term record analysis “Ultra Low Frequency” (ULF) (more
than 0.003 Htz) component is also singled out.
Experience of the Russian investigations and their results that have been car-
ried out by many foreign authors, show that these recommendations should be cor-
rected. It’s mostly relevant to VLF range. This is the following corrected scheme of
frequency ranges for spectral analysis of HRV:
Names of spectrum components Frequency range, Htz Period, seconds
HF 0,4 – 0,15 2,5 – 6,6LF 0,15 – 0,04 6,6 – 25,0
VLF 0,04 – 0,015 25,0 – 66,0ULF Less than 0,015 More than 66,0
29
The offered limitation for VLF range to 0.015 Htz is explained by the fact
that making analysis of 5-minute long records we, in fact, can accurately identify
only oscillations with a period that is 3-4 times less than signals registration period
(about 1 minute). That is why all oscillations with the period more than one minute
are offered to be referred to as of ULF range and to single out corresponding sub-
ranges in it.
As for the spectral analysis the absolute summary power of a range, average
power of a range, maximum harmonics value and relative value (in percent of Total
Power – TP) are usually calculated for each component. So TP is the sum of range
power HF, LF and VLF. According to the spectral analysis data (CR analysis) the
following indicators are calculated: Index of centralization (IC=(HF+LF)/VLF)
and Index of vagosympathetic interaction LF/HF.
Pic.6. Typical HRV spectrogram by using the method of Fure’s quick transfor-mation (FQT).
5.6. Other methods of HRV analysis.
Digital filtration. Methods of digital filtration are appointed for fast analysis
of brief parts of ECG (less than 5 minutes) and they allow making quantitative eval-
uation of HRV periodical components. For instance, this is sliding averaging on the
definite number of consecutive cardiointervals. For identifying of the slow waves of
the 1-st degree averaging on 5 or 9 cardiointervals is applied.
30
Nonlinear dynamics methods. Many influences on HRV including neurohu-
moral mechanisms of higher vegetative centers cause nonlinear character of changes
of CR for the description of what application of special methods is needed. To this
problem has been recently paid much attention abroad as well as in our country. For
the description of nonlinear properties of variability section of Puankare, chest spec-
tral analysis, attracter’s graphics, singular decomposition, Lapunov’s exponent, Kol-
mogorov’s entropy were applied. All these methods are recently only of investiga-
tion interest and their practical application is limited. Along with this it is necessary
to mention method of functional state estimation based on the chaos’ theory used in
“Neurosoft”, “Vita-Rhythm” device. In2001 the special symposium under the name
“Theoretical and Applied Aspects of Nonlinear Chaos and Fractal’s Dynamics in
Physiology and Medicine” in Novokuznezk was held.
6. Reproductiveness and comparativeness of data.
Constantly acting regulatory mechanisms provide adequate adaptation re-
sponses of an organism caused by continuous environmental changes. It means that
the functional state of different regulation links is constantly changing and it is im-
possible to obtain identical results after repeated HRV investigations.
That is why reproductiveness of data can not be equal to 100% (completely
the same). High reproductiveness means only qualitative but not quantitative corre-
spondence of two compared records (registrations) obtained from one and the same
person even after rather small period of time. Discussing the problem of reproduc-
tiveness of HRV analysis data one must consider high sensitivity of the vegetative
nervous system for external and internal influences, typical peculiarities of patient
and his health.
In a number of cases (initial stages of some diseases, non-stability of vegeta-
tive regulation) high reproductiveness should not be expected at all. Daily changes
in vegetative regulations must also be considered. For the high data (of HRV analy-
sis) reproductiveness obtaining it is recommended to keep to registration’s method
described in chapter 4.2.
31
Comparativeness of registrations and HRV analysis results means that it is
possible to compare data obtained in different clinics and establishments with the
help various equipment and programs. Without the possibility of such comparison it
is impossible for methods of HRV analysis to develop in future. It is a question of
comparing main (key) indicators of statistical and spectral analysis.
Clinic physiological interpretation of these indicators and creating new evalu-
ation algorithms based on them ought to be the subject of further scientific investi-
gations. However, if the min indicators of HRV are essentially different depending
on the types of applied devices and programs, it will be of no progress in HRV anal-
ysis.
Present recommendations for application of different ECG systems for HRV
analysis envisage using of special testing system that must include set of control
files, special test-program and special standard ECG database. All firmware com-
plexes manufactured in Russia must pass testing procedure to determine the accor-
dance to the accustomed HRV analysis standards.
As a standard test-system it is recommended to use “HRV-test” complex that
includes set of real and generated signals (ECG) and also the results of data pro-
cessing by standard analysis program (for HRV).
Three levels of testing are studied:
1. Testing of the system that fulfils the functions of identifying of R-cogs, R-R
intervals duration, measuring, formation of normalized cardiointervals row
and calculating of key indicators of HRV.
2. Testing of the system that has only functions of making normal range of car-
diointervals and calculating key (standard) indicators of HRV.
3. Testing of the system that is only for calculating key (standard) indicators of
HRV.
Adjusting of such levels of testing it is necessary for making standards of not
only complete firmware complexes but also of special programs products designed
32
for HRV analysis included into serial produced devices as well as designed for deal-
ing with database or separately gathered R-R intervals files autonomously.
The list of recommended set of HRV indicators is given in 1-st supplement
and formulas for calculating them are given in the 2-d one.
7. Evaluation of HRV analysis results.
Clinical interpretation (and physiological) of results obtained is of the
paramount importance for investigators and clinicians who use HRV analysis. How-
ever, nowadays there is no consentaneous opinion concerning HRV analysis results
interpretation. But there are definite clinical physiological evaluations that are al-
ready more or less alike in the major part of publications. For some indicators there
are original but yet disputed interpretations that need more details.
In the present chapter the materials on HRV analysis results evaluation are
given only the main indicators that are used in Russia more often are listed; their
clinical physiological interpretation based on the traditional viewpoint concerning
vegetative regulation of heart, sympathetic and parasympathetic influence, and role
of subcortical cardio-vascular centre and higher physiological control centre in it.
Much attention is paid to the complex evaluation of the organism functional
conditions according to RSAI (Regulatory System Activity Indicator).
An important role belongs to the comparing of the obtained during results
evaluation data with standard indicators.
Standard considered to be some complex of statistical data obtained while ex-
amination of deliberately selected group of healthy people needs to be clarified as
the HRV analysis application.
As it is not a question of relatively stable parameters evaluation but of rather
changeable parameters of vegetative regulation so it would be more correct to re-
gard standard being the functional optimum.
Here it is to be pointed out that individual organism optimum does not always
coincide with average statistical standard for common adaptation reactions take
place differently depending on the conditions that surround the examined person and
33
depending on his personal functional reserves. Notion of the physiological standard
has been made out in space medicine. It indicates at reservation of sufficient level of
organism’ functional abilities. So that the homeostasis of the main systems of the or-
ganism is provided with minimum of regulatory mechanisms stress. So the value of
the biggest part of indicators of HRV must not exceed definite levels corresponding
to the definite age, sexual, professional and regional group. This point is mostly ful-
filled by application of complex HRV analysis results evaluation (see below). There
is also a notion of clinical standard that characterizes indicators values in visually
healthy persons. However, as it is known, nosologic approach is based on the evalu-
ation changes, mainly on structural, metabolic and energic metabolic levels of orga-
nization of living system, and it also takes into account the state of regulatory sys-
tems. Thus, norm problem with reference to an estimation of HRV requires the fur-
ther profound development.
It is necessary to point, that the materials of the given unit carry only recom-
mendatory character. They can be especially useful for the beginning experts for the
correct use of a method and understanding of its opportunities.
7.1. Indicators of the statistical analysis (temporarily analysis)
Average square-law deviation (ASD, SD). The calculation of ASD is the
simplest procedure of the HRV statistical analysis. The meanings of ASD are ex-
pressed in milliseconds (ms). The normal meanings of ASD are within the limits of
40 - 80 ms. However these meanings have age-sexual features, which should be
taken into account at an estimation of results of research.
The increase or decrease of ASD can be connected both with an independent
circle of regulation, and with central (both with sympathetical and with parasym-
pathetical influences on a rhythm of heart). At the analysis of short records, as a
rule, increase of ASD points to intensifying of an independent regulation, that is the
body height of influence of respiration on a rhythm of heart, that is observed in
dream more often.
34
The decrease of ASD is connected with intensifying of a sympathetic regula-
tion, which inhibits the activity of an independent circle. The sharp drop of ASD is
caused by considerable strain of regulation systems, when highest levels of control
are included in the process of regulation that conducts to almost complete inhibition
of activity of an independent circle. The information on physiological sense similar
to ASD can be received from a parameter of cooperative capacity of a spectrum -
ТР. This parameter differs by that it characterizes only periodic processes in a
rhythm of heart and does not contain a so-called fractional part of process, that is
nonlinear and acyclic components.
RMSSD is a parameter of activity of a parasympathetical link of vegetative
regulation. This parameter is calculated on dynamic series of differences of mean-
ings of consecutive couples of cardiointervals and does not contain slow-wave com-
ponents of CR. It represents activity of an independent contour of a regulation. The
higher is the meaning of RMSSD, the more active is the part parasympathetic regu-
lation. In norm meanings of this parameter are within the limits of 20-50 ms. The
similar information can be received from a parameter pNN50, which expresses in
persentage the number of differential meanings more than 50 ms.
The index of pressure of control systems (IP) characterizes activity of
mechanisms sympathetic of a regulation, state of the central circle of a regulation.
This parameter is calculated on the basis of the analysis of the diagram of cardioint-
erval distribution - variational pulsegramm. The activation of the central circle, in-
tensifying of sypmathical regulation during mental or phisical exercises is shown by
stabilization of a rhythm, decrease of disorder of the length of cardiointervals, by in-
creasing of quantity of intervals with the same duration (the increas of АМо). The
form of histograms changes, there is their narrowing with simultaneous raise in
height.
Quantitatively it can be expressed by the ratio of height of a histogram to its
width (see above). This parameter has received the name of an index of a control
systems pressure (IP). In norm IP changes within the limits of 80-150 standard
35
units. This parameter is extremely sensitive to intensifying of sympathetic nervous
system tone. The small load (physical or emotional) enlarges IP in 1.5-2 times. At
considerable loads it grows at 5-10 times. Patients with a constant tension of circle
systems have IP in rest equal to 400-600 standard units. IP in rest of patients with
attacks of a stenocardia and myocardial infarction reaches 1000 - 1500 units.
7.2. Parameters of a spectral analysis (frequency analysis)
Capacity of high-frequency component of a spectrum (respiratory
waves). Activity of sympathetical department of vegetative nervous system, as one
of components of vegetative balance, can be estimated on a degree of activity inhib-
ition of an independent circle of a regulation, for which parasympathetic section is
responsible.
Vagus activity is a basic amounting of HF component. It is well reflected by a
parameter of capacity of СR respiratory waves in absolute digits and as relative
quantity (in % from cooperative capacity of a spectrum).
Usually respiratory amounting (HF) makes 15-25 % of cooperative capacity
of a spectrum. The drop of this share to 8-10 of % specifies shift of vegetative bal-
ance to the side of prevalence of a sympathetic department. If the quantity does not
fall below 2-3 % than it is possible to speak about sharp prevalence of sympathetic
activity. In this case parameters RMSSD and pNN50 also decrease essentially.
Capacity of allow-frequency amounting of spectrum (slow waves 1st or-
der or vasculomotor waves). This parameter (LF) characterizes a condition of a
sympathetic department of vegetative nervous system, in particular, systems of a
regulation vascular tone. In norm the sensitive receptors of a sinocarotid zone per-
ceive changes of arterial pressure value and afferent nervous impulsation acts in
vasculomotor centre of an oblong brain. Afferent synthesis (processing and analysis
of the acting information) is carried out here and the signals of management goes
into a vascular system (afferent nervous impulsation). This process of the vascular
tonus control with a feedback on lubricous muscular fibre of vessels is carried out
by vasculomotor centre constantly. The time necessary for vasculomotor centre to
36
operate receptions, to processe and transfer of the information changes from 7 up to
20 sec.; usually it is equal to 10 -12 sec. Therefore in a rhythm of heart it is possible
to find out waves with frequency close to 0,1 Htz (10), which have received the
name vasculomotor. For the first time these waves were observed by Mayer with
the co-authors (1931) and therefore they are sometimes referred to as Mayer waves.
The capacity of slow waves of 1-st order determines activity of vasculomotor
centre.
The transition from a position "laying" in positions "standing" conducts to
substantial growth of capacity in this range of fluctuations CR. The activity of vas-
culomotor centre falls with age and the persons of elderly age have this effect prac-
tically absent (see pic. 7). Instead of slow waves of 1-st order, the capacity of slow
waves of 2d order is enlarged. It means, that the process of a AD regulation is car-
ried out at participation of nonspecific mechanisms by activization of the sympathet-
ic department of a vegetative system. Usually in norm the percentage share of vas-
culomotor waves in a "laying" position is from15 till 35-40%.
It is necessary to mention also a parameter of dominant frequency in a range
of vasculomotor waves. Usually it is within the limits of 10-12 sec. Its increase to
13-14 sec can specify the drop of activity of vasculomotor centre or on retardation
of baro-reflector regulation.
Capacity of a "very" low-frequency amounting of a spectrum (slow
waves of 2d order). Spectral amounting of a cardiac rhythm in a range 0.05 – 0.015
Htz (20 - 70 sec.), in opinion of many foreign authors, characterizes activity of a
sympathetic department of vegetative nervous system. However in this case it is
about more complex influences from the side of above-segmental level of regulation,
as the VLF amplitude is closely connected with psychoemotional state and function-
al condition of a cortex of a brain. It is shown, that VLF reflects cerebral ergotropic
influence on below levels and allows to judge a functional condition of a brain at a
psychogenic and organic pathology of a brain (N.B.Haspekowa, 1996).
37
The purposeful researches hold by A.N.Fleishmann (1999) have shown the im-
portant meaning of the HRV and VLF - range analysis. In the classification of ВСР
spectral components, offered by him, the ratio of HF, LF and VLF amplitudes is taken
into account, and 6 classes of spectrograms (see pic. 8) are surveyed. It is also
shown by A.N.Fleishmann that the capacity of VLF-fluctuations of HRV is sensitive in-
dicator of management of metabolic processes and well reflects deficit energy states.
As this approach has no foreign analogues it is expediently to present its more detailed
description.
In a pic. 9 the circuit of an estimation of deficit energy states with use of a series
of functional tests (account in mind and hyperventilation) is submitted. High in compar-
ison with norm VLF level can possibly be treated as hyperadaptiv state, the reduced
VLF level specifies on энергодефицитное state.
Despite of conditional and in many respects still disputable character of such in-
terpreting of VLF changes it can be useful for research of both healthy people, and pa-
tients with various states connected with infringement of metabolic and power pro-
cesses in an organism.
The mobilization of energetic and metabolic reserves at functional influences
can be reflected by changes of capacity of a spectrum in VLF - range. At increasing of
VLF in reply to a load it is possible to speak about hyperadaptive reaction, at its drop -
about afterload deficit energy.
Thus, VLF characterizes influence of the highest vegetative centres on cardi-ovascular subcortical centre, it reflects a condition of neurohumoral and metabolic levels of a regulation. VLF can be used as a reliable marker of blood circulation con-nection degree between autonomous (segmental) and supersegmental regulation lev-els, including hypophysis-hypothalamus and cortical levels. To be normal VLF=15-30% of total spectrum power.
400
600
800
1000
1200мсек
стоя сидя Nлежа
38
А.
400
600
800
1000
1200
N
мсек
стоя сидялежа
Б
Pic.7. Cardiointervalogram by active orthostasis test with young (at the top) and old (at the bottom) patients.
39
Pic.8. Kinds of spectrum characteristics of cardiac rhythm (by A.N.Fleishmann, 1999).
40
Classification of energo-changed states on the basis of dynamics of SPM spec-trums MCG on the load.
(1 – background; 2 – count in wit; 3 – recovering 1; 4 – hyperventilation; 5 – re-covering 2)
Pic 9. Scheme of evaluation of deficit energy states (by A.N.Fleishmann, 1999).
41
7.3Functional condition integrated assessment
Integrated assessment of heart rhythm variability contem-plates functional condition (not diseases!) diagnostics. Vegetative balance changes in the form of sympathetic section activation rates as nonspecific adaptation reaction component in response to different stressful influences.
One of the ways of such reactions estimation is the regulatory systems ac-tivity index (RSAI) computation. It is calculated in numbers with special algorithm that takes account of statistic figures, histogram ones and cardiointervals spectrum analysis data.
RSAI allows to differentiate regulatory systems different tension degrees and appreciate organism adaptation abilities (R. M. Baevsky, 1979). RSAI calcula-tion is realized by a program that accounts next five criterions:
A. Regulation total effect by pulse frequency (PF) figures.B. Total regulatory mechanisms activity by standard deviation – SD (or by total
spectrum power - SP)C. Vegetative balance by a complex of figures: RMSSD, HF, IC.D. Activity of vasomotor centre, that regulates vascular tone, by the first-order
slow waves (LF) spectrum powerE. Cardiovascular subcortical nerve centre or supersgmental regulation levels
activity by the second-order slow waves (VLF) spectrum power. RSAI value evaluates in 10 to 10. Basing RSAI values analysis next func-
tional states may be diagnosed:1. Optimal (working) regulatory systems exertion condition, necessary for
the active organism and environment balance support (RSAI in norm=1-2).
2. Moderate regulatory systems exertion condition, when for the adapta-tion to the environment conditions organism requires additional func-tional reserves. Such conditions arises in the processes of the adapta-tion to a work activity, under emotional stresses or under negative eco-logical factors influence (RSAI=3-4)
3. Marked regulatory systems exertion condition, that relates to active de-fensives mobilization, including sympathetic-adrenal and hypophysis-adrenal systems activity rise (RSAI=4-6)
4. Regulatory systems overstrain condition, for which it is typical protec-tive-adaptation mechanism insufficiency, their failure to provide ade-quate organism reaction to environmental impact. Here excess regula-tory systems activation doesn’t fortified by existent functional reserves (RSAI=6-7)
5. Regulatory systems exhaustion (asthenisation) condition, where man-agement mechanisms activity decreases (regulation mechanisms insuf-
42
ficiency) and internals of pathology appear. At this point specific changes distinctly prevail on nonspecific (RSAI=7-8).
6. Regulatory systems “brake” condition (adaptation failure), when spe-cific pathological changes dominate and adaptation self-regulating mechanisms partly or fully violated (RSAI=8-10)
When RSAI value is estimated, one can conditionally distinguish three functional conditions zones, for clearness represented like the “traffic lights”: GREEN means that everything is in norm, and doesn’t call for any special actions by prophylactic and treatment. YELLOW – indicates prophylactic and treatment ac-tions necessity. Finally, RED shows, whether at first diagnostics and then the treatment of possible diseases is required. The isolation of the green, yellow and red zones of health makes it possible to characterize the functional state of man from the point of view of the risk of the devel-opment of disease. For each step of the “stairs of states” the “diagnosis” of functional state is provided according to the degree of the manifestation of the regulatory systems voltage. Furthermore, there is a possibility to refer the examined patient to the one of 4 functional states according to the classification accepted in prenosological diagnostics (R.M.Bayevskiy, A.P.Beresneva, 1997):
• The normal state or the state of satisfactory adaptation (IARS=1-3),• The state of functional stress (IARS=4-5),• The state of overstress or the state of unsatisfactory adaptation (IARS=6-7),• The state of the exhaustion of regulator systems or the disruption of adaptation
(IARS=8-10)
Developed by IVNMT “Ramena” the complex “Varikard” not only makes it possi-ble to calculate IARS and to evaluate functional state, but also forms the individual con-clusions (see Picture 10). It is necessary to note that IARS does not have analogs in foreign studies. The disquality of IARS is that it makes its possible to obtain only the discrete estimations of functional states, which is insufficient during the dynamic con-trol. For guaranteeing the continuous scale of estimations mathematical models can be used as quantitative dependences between the collection of numerical signs (values of HRV indices) and the functional states of organism (Bayevskiy R.M., Semenov YU.N., Chernikov A.G., 2000).
43
Pic.10. Sample of complex conclusion according to the results of HRV analysis (firmware complex “Varikard”).
7.4. The evaluation of HRV analysis results in functional trials conducting The HRV analysis estimation in functional load trials conducting requires a spe-
cial attention. The development of separate medical instructions for each functional trial is necessary. The most complete information about the HRV analysis in different func-tional trials conducting is contained in V.M.Mikhaylov monograph (2000). Some general
44
recommendations regarding the interpretation of HRV indices in the functional trials consist of the following:
1. The functional state of organism estimation (vegetative balance, the degree of the stress of regulatory systems etc.) has the greatest value in the initial period (back-ground) prior to the beginning of functional action. The interpretation of data in the dif-ferent stages of functional trial must be carried out, first of all, via comparison with the initial state.
2. In all functional trials there is a transient process between the initial state and the new functional state, which is formed in the process of conducting the trial. This transitional process has different nature and different duration in different functional tri-als. The isolation of transitional process from the common record and its estimation by special methods is one of the important problems of functional testing. The most valu-able information about the state of regulator mechanisms is frequently contained in the transitional process. The methods of the transitional processes analysis are not includ-ed in the given systematic recommendations.
3. Under the effect of functional action the new functional state is formed, and it is not steady.
This fact must be considered especially, while analyzing the dynamics of HRV indices, which reflect the subtle interconnections between different links of regulator mechanism. Therefore it is necessary to separate different stages of functional trial for the estimation.
4. At least two stages of the functional trial should be distinguished: the stage (or period) of direct effect of the corresponding factor on the organism and the stage (or period) of restoration. There is a transitional process between the end of action and the beginning of restoration also, which requires recognition, isolation and special estima-tion.
5. During the HRV indices estimation in the different stages of functional trial it is recommended to evaluate not only their average values, but also dynamics of changes, and the synchronization of these changes.
CONCLUSION.BASIC TRENDS IN FURTHER DEVELOPMENT OF THE HRV ANALYSIS
METHODS On present stage of the practical use of HRV analysis methods in the applied
physiology and clinical medicine the approaches to the physiological and clinical inter-pretation of above represented data make it possible to effectively solve many prob-lems of diagnostic and prognostic profile, evaluation of functional states, control of the effectiveness of therapeutic and prophylactic actions, etc. However, the possibilities of this methodology are far from exhaustion and its development continues. The brief enu-meration of some trends in further development of the HRV analysis methods, which are developed mainly in Russia Is given below. Their number includes:
• Study of the slow waves of 2nd order (VLF) and ultra-slow-wave components of the heart rate spectrum (ULF) - oscillations at frequencies below 0,01 Hz (100 s), including the minute and hour waves (ultraDian rhythms).
• Development of the methodology of variation pulsimetry, including differential chronocardiography and new approaches to the statistical analysis of the heart rate variability (Fedorov V.F., Smirnov A.V., 2000).
• Use of heart rate variability for evaluating the level of stress, degree of the ten-sion of regulator systems (Computer electrocardiography, M., 1999).
45
• Study of the heart rate variability in children and adolescents, including the influ-ence of school loads and the age-sexual aspects (Bezrukich M.M., 1981, Shlyk N.I., 1991).
• Use of methods of the heart rate variability analysis in aerospace biomedicine, in medicine of extreme actions and in different fields of applied physiology (Grigo-ryev A.I., Bayevskiy R.M., 2001).
• Development of the clinical directions of the use of the method: A) in the surgery - the control of anesthesia, b) in neurology - the differential estimation of mor-phological and functional damages, c) in oncology - attempt at the estimation of the degree of metabolic disturbances (Computer electrocardiography, 1999, Fleishmann A.N.1999).
• Development of the new principles of heart rate variability analysis use in the cardiological clinic - the estimation of pathologic process gravity, the prognosti-cation of outcomes and effectiveness in the treatment, the estimation of gravity and risk with arrhythmias (Dovgalevskiy P. 0., O.K.Rybak 1996, Ivanov G.G., etc.1999, Minakov E.V., etc.1998, Mironov V.A., 1998, Yavelov I.S., etc., 1997, Smetnev A.S., etc., 1995).
In conclusion it should be again emphasized that in the given systematic rec-ommendations only the aspects of the use of the so-called “short-term” records of heart rate (from several minutes to several hours) were examined. Research methodology and analysis principles of such records differ essentially from methods of approach of higher complexity in case of working with 24-hour HRV records which can be obtained due to Holter’s monitoring. Undoubtedly, day observe data let us evaluate mechanisms of neuroendocrine regulation condition more deeply and our native researchers have achieved a great success in this field (G.V.Rabyki-na, A.V.Sobolev, 1998; V.M.Makarov, 1999).
However, 24-hours analysis is much more valuable and difficult, HRV day records analysis is not worked out enough, particularly it concerns to the transitional processes. Short records incontestable advantages are wider range of method us-ing, simplicity of apparatus and software, possibility of fast results obtaining. All this determines wide perspective prevalence of HRV analysis methods in applied physi-ology, prophylactic medicine and clinical practice.
LITERATURE
Анохин П.К. Принципиальные вопросы общей теории функциональных систем. Принципы системной организации функций. М., Наука, 1973, С.5-61.
Баевский Р.М. К проблеме прогнозирования функционального состояния челове-ка в условиях длительного космического полета. Физиол. Журн. СССР,1972, 6, с.819-827.
Баевский Р.М. Кибернетический анализ процессов упраления сердечным ритмом. Актуальные проблемы физиологии и патологии кровообращения. М., Медицина.1976. С. 161 -175.
Баевский Р.М., Кириллов О.И., Клецкин С.З. Математический анализ изменений сердечного ритма при стресса. М., Наука, 1984.С.220
46
Баевский Р.М., Берсенева А.П. Оценка адаптационных возможностей организма и риск развития заболеваний. М., Медицина. 1997. С. 265.
Баевский Р.М. Прогнозирование состояний на грани нормы и патологии. М., Ме-дицина, 1979, 205 с.
Баевский Р.М., Семенов Ю.Н., Черникова А.Г. Анализ вариабельности сердечного ритма с помощью комплекса “Варикард” и проблема распознавания функциональных со-стояний. Хронобиологические аспекты артериальной гипертензии в практике врачебно-летной экспертизы (Разсолов Н.А., Колесниченко О.Ю.), М.. 2000.С. 167 –178
Баевский Р.М., Иванов Г.Г. Вариабельность сердечного ритма: теоретические ас-пекты и возможности клинического применения. Ультразвуковая и функциональная диа-гностика. 2001, 3, с.106 -127
Безруких М.М. Регуляция хронотропной функции у школьников 1-4 классов в процессе учебных занятий. Возрастные особенности физиологических систем у детей и подростков. М.,1981. С.249-254.
Воробьев В.И. Исследование математико-статистических характеристик сердечно-го ритма как метод оценки реакции лиц разного возраста на мышечную нагрузку. Дисс. канд. биолог. наке, М., ИМБП. 1978. 178 с.
Вариабельность сердечного ритма. Теоретические аспекты и практическое применение. Тезисы международного симпозиума 12-14 сентября 1996 г.. Ижевск. 1996. С.225
Власов Ю.А., Яшков В.Г., Якименко А.В. и др. Метод последовательного парного анализа ритма сердца по интервалам RR. Радиоэлектроника, физика и математика в био-логии и медицине. Новосибирск. 1971. С.9-14.
Воскресенский А.Д,, Вентцель М.Д. Статистический анализ сердечного ритма и показателей гемодинамики в физиологических исследованиях. М., Наука, 1974, 221 с.
Габинский Я.Л. Вариационная пульсометрия и автокорреляционный анализ в оценке экстракардиальной регуляции сердечного ритма. Автореф. Дисс. Канд. мед. Наук. Свердл. Мед. Ин-т.,1982, 22 с.
Гаврилушкин А.П.,Маслюк А.П. Теоретические и практические аспекты нелинейных хаотических колебаний ритма сердца, Медленные колебательные процессы в организме человека. Теоретические и прикладные аспекты нелинейной динамики, хаоса и фракталов в физиологии и медицине. Материалы 3-го Всероссийского симпозиума 21-25 мая 2001 г. Новокузнецк, 2001
с. 37-48
Григорьев А.И., Баевский Р.М. Концепция здоровья и проблема нормы в космиче-ской медицине. М., Слово, 2001, 96 с.
Довгалевский П.Я., Рыбак О.К. Возможность использования системного анализа в оценке нейрогуморальной регуляции сердечного ритма у больных ИБС. Международный
47
симпозиум «Вариабельность сердечного ритма. Теоретические аспекты и практическое применение», Ижевск, 1996, с.29-30
Жемайтите Д.И. Ритмичность импульсов синоаурикулярного узла в покое и при ишемической болезни сердца. Автореф. дисс.. канд.мед. наук. Каунас, Мед. Ин-т, 1965, 51 с.
Жемайтите Д.И. Возможности клинического применения и автоматического ана-лиза ритмограммю Дисс. докт. мед. наук. Каунас. Мед.ин-т. 1972. 285 с.
Иванов Г.Г., Дворников В.Е., Баев В.В. Внезапная сердечная смерть: основные ме-ханизмы, принципы прогноза и профилактики. Вестник РУДН. 1998, N1,144-159.
Клецкин С.З. Проблема контроля и оценки операционного стресса ( на основе анализа ритма сердца с помощью ЭВМ). Дисс. докт. мед наук. М., Ин-т серд.сосуд.хирург. АМН СССР, М., 1981. 298 с.
Компьютерная электрокардиография на рубеже столетий. Международный симпо-зиум. Москва 27-30 апреля 1999 г. Тезисы докладов. М., 1999. С.320
Кудрявцева В.И. К проблеме прогнозирования умственного утомления при дли-тельной монотонной работе. Автореф. дисс. канд. биол. Наук. М., ИМБП, 1974, 23 с.
Макаров Л.М. Холтеровское мониторирование . М., Медицина, 2000, 104 с.
Математические методы анализа сердечного ритма. Материалы 1-го Всесоюзного симпозиума. Под ред. Парина В.В. и Баевского Р.М.. М., Наука, 1968
Медленные колебательные процессы в организме человека: Теория и практиче-ское применение в клинической медицине и профилактике. Сборник научных трудов сим-позиума 27-29 мая 1997 г., Новокузнецк, 1997.С. 194.
Минаков Э.В., Соболев Ю.А, Стрелецкая Г.Н., Минакова Н.Э. Использование ма-тематического анализа сердечного ритма в процессе реабилитации больных гипертониче-ской болезнью. Международный симпозиум «Вариабельность сердечного ритма. Теоре-тические аспекты и практическое применение», Ижевск, 1996, с.42-43
Михайлов В.М. Вариабельность сердечного ритма. Опыт практического примене-ния. Иваново, 2000, 200 с.
Миронов В.А. Клинический анализ волновой структуры синусового ритма сердца при гипертонической болезни. Автореф. дисс. докт.мед.наук., Оренбург, 1998, 53 с.
Миронова Т.В., Миронов В.А. Клинический анализ волновой структцры синусо-вого ритма сердца ( Введение в ритмокардиографию и атлас ритмокардиограмм). Челя-бинск, 1998. С.162.
Нидеккер И.Г. Выявление скрытых периодичностей методом спектрального анали-за. Дисс. канд.физ-мат. наук. М., ВЦ АН СССР. 1968. 131 с.
48
Никулина Г.А. Исследование статистических характеристик сердечного ритма как метод оценки функционального состояния организма при экстремальных воздействиях. Автореф. дисс. Канд. мед. наук . М., ИМБП, 1974, 30 с.
Парин В.В., Баевский Р.М. Введение в медицинскую кибернетику. М., Медицина, 1966, С.220.
Парин В.В., Баевский Р.М., Волков Ю.Н., Газенко О.Г. Космическая кардиология. Л., Медицина, 1967. С.206
Рябыкина Г.В., Соболев А.В. Анализ вариабельности ритма сердца. Кардиология, 1996, 10, с.87 -97
Рябыкина Г.В., Соболев А.В. Вариабельность ритма сердца. М., Из-во "СтарКо", 1998.
Селье Г. Очерки об адаптационном синдроме. Пер. с англ. М., Медгиз, 1960, С.275.
Сметнев А.С., Жаринов О.И., Чубучный В.Н. Вариабельность ритма сердца, желудочковые аритмии и риск внезапной смерти. Кардиология, 1995, 4, с.49-51
Федоров В.Ф., Смирнов А.В. О некоторых неиспользованных возможностях статистических методов в кардиологии. Клинические и физиологические аспекты ортостатических расстройств» М., 2000, с.138-148
Флейшман А.Н. Медленные колебания гемодинамики. Новосибирск, 1999.С.264.
Флейшман А.Н. Медленные колебания кардиоритма и феномены нелинейной динамики: классификация фазовых портретов, показателей энергетики, спектрального и детрентного анализа. Медленные колебательные процессы в организме человека. Теоретические и прикладные аспекты нелинейной динамики, хаоса и фракталов в физиологии и медицине. Материалы 3-го Всероссийского симпозиума 21-25 мая 2001 г. Новокузнецк, 2001
с.49 –61
Хаспекова Н. Б. Регуляция вариативности ритма сердца у здоровых и больных с психогенной и органической патологией мозга. Дисс. докт.мед.наук. М., Ин-т ВНД.1996. 236 с.
Хаютин В.М., Лукошкова Е.В. Спектральный анализ колебаний частоты сердцеби-ений: физиологические основы и осложняющие его явления. Российский физиол. Журн. Им. И.М. Сеченова, 1999,85 (7),с.893-909
Шлык Н.И. Сердечный ритм и центральная гемодинамика при физической актив-ности у детей. Ижевск, 1991. С 417.
Goldberger A.Is the normal heartbeat chaotic or homeostatic? News in Physiological Sci-ences, 1991:6:87-91.
Heart rate variability. Standatds of Measurement, Physiological interpretation and clinical use. Circulation, 1996 ,V.93, P.1043-1065
49
Parin V.V., Baevsky R.M., Gazenko O.G. Heart and circulation under space conditions. Cor et Vasa, 1965, 7 (3), p.165-184
Supplement 1
List of main indices of heart rate variability
№ Abbreviation Names of Indices Physiological Interpretation
1 PF Pulse frequency Middle level of functional systemcircula-tion
2 SDNN Standard deviation of full sol-id cardiointervals
Total effect of vegetative regulation circu-lation
3 RMSSDSquare root from sum of dif-ference of logical row of car-diointervals
Activity of parasympathetic chain of vege-tative regulation
4 pNN50
Number of pairs of cardioint-ervals with difference more than 50 ms in percentage to the total numbers of them in massive
Index of prevalence level of parasympa-thetic chain regulation over sympathetic (relative value)
5 CV Coefficient variation of full solid of cardiointervals Standard index of regulation total effect
6 MxDMn Difference between max and min values of cardiointervals Maximal amplitude of regulator influences
7 Mo Mode More presumable level of cardiovascular function system
8 AMo Amplitude mode Nominal index of activity of sympathetic chain regulation
9 SI Stress index (index of regula-tory systems tension)
Level of tension of regulatory systems (lev-el of activity prevalence of central mecha-nisms regulation above autonomic)
10 CC1 Value of first coefficient of function autocorrelation
Level of activity of autonomic circle regu-lation
11 CC0
Amount of shifts of function autocorrelation before obtain-ing coefficient correlation value less than 0
Level of activity of central circle regulation
12 TP Total power of HRV spec-trum in ms2
Total absolute level of regulatory systems activity
13 HF, (%)
Power spectrum of high fre-quency component of vari-ability in percentage from to-tal power waves
Relative level of activity of parasympathet-ic chain regulation
14 LF, (%) Power spectrum of low fre-quency component of vari-ability in percentage from to-
Relative level of activity of vasomotor cen-tre
50
tal power waves
15 VLF, (%)
Power spectrum of very low frequency component of vari-ability in percentage from to-tal power waves
Relative level of activity of sympathetic chain regulation
16 HFavMiddle power value spectrum of high frequency component HRV in ms2
Middle absolute level of activity of parasympathetic chain vegetative regula-tion
17 LFavMiddle power value spectrum of low frequency component HRV in ms2
Middle absolute level of activity of vaso-motor centre
18 VLFavMiddle power value spectrum of very low frequency compo-nent HRV in ms2
Middle level of activity of sympathetic chain vegetative regulation (mostly above segmental sections)
19 (LF/HF)avDifference of middle values of low and high frequency com-ponent HRV
Relative activity of subcortical sympathetic nervous centre
20 IC Index o centralizationLevel of centralization of cardiac rhythm managing (prevalence of activity of central circle regulation above autonomic)
The above-presented list of indices does not delete of using other methods of analysis and development.
Supplement 2.Accounting formulas for calculation of main indices of heart rate variability.There are three forms of data presentation of HRV mathematical analysis:
1. dynamic row of NN intervals – Nni, i=1,2….n;2. data calculated on the basis of difference between NN-intervals;3. new row of discrete values xi, i=1,2….N. The scheme of a new row is based on
the state, that HRV is set by unceasing function from time – x(t), defined out of number of elementary actions, i.e. moments of appearance of R-notches. Func-tion values in these moments are equal to values corresponding to NN-intervals. Function values in gaps of time between appearance of R-notches are calculat-ed by the method of splin cubical interpolation. Row is built by method of quan-tum function x(t) with step 250 ms.Statistic methods.The calculation of basic parameters of variability must include the following in-
dices:HR is calculated as quantity of NN-intervals in record, sharing at duration of their
record:
∑=
⋅⋅= n
ii мсNN
nHR
1)(
100060 (in 1min); (1)
Middle value:
51
∑=
=N
iix
Nx
1
1 (ms), (2)
Where xi is a value of i-quantum element of function x(t), i=1,2,...,N;Dispersion is meant to its selectional (empirical) value and calculated by formu-
la:
2
111 )x(x
)(N-D
N
i=i −= ∑ (ms2); (3)
Middle solid deviation (SDNN) or σ is defined as a share root out of dispersion:
Dσ = (ms); (4)variation coefficient (CV) is inter-changed of its empirical characteristics and
counted as relation (in percents) of average root deviation to appropriate mathemat-ical waiting
(5)
RMSSD – root mean sum successful deviation counted as
(ms)
(6)
PNN50 – relation (in per cents) of NN-intervals, difference of those (xi–xi-1)> 50 ms, to total number of NN-intervals.
1. Geometric methods
Geometric methods are based on drawing graphic of function of density of variability distribution x(t). Graphic drown on next algorithm:1. The scheme of histogram with 50ms stepping from 0.3 to 1.7 s. Thus we get 28
ranges of function. We get ordinates of histogram ranges by relation of xi ele-ments quantity which come in range of total N-elements quantity and have meas-ure of 1/50 (ms)
2. The scheme of the 2nd histogram quantity of ranges (classes) in which we get with Shtrurges's rule:
52
RMSSD1
n 1−( )1
n 1−
i
NNi NNi 1+−( ) 2∑=
⋅
CVσx
100%⋅
Nclass Int 1 1.44 ln N( )+( )
(7)
where N is total number of elements xi, and int is a whole part selection func-tion.
3. Variation amplitude (-∆x) estimated (MxDmn in first approach) by empir-ical formula:
(8)
4. In each class of 2nd diagram counted the number of discrete xi indices. On graphic for each class plotted a dot with abscise equal to average value of all xi, and ordinate equal to quantity of discrete xi indices in this class. If there is no data in this class dot is absent.
5. Ordinates of dots in 2nd histogram multiplied on normalization coefficient:
(ms)
(9)
and connected with smooth curve by spline interpolation method. The 2nd
graphic has the same measurement as the 1st one.
In function of variability density distribution graphic we can get to know next indices:
Moda amplitude (Amo) correlates with the level of maximum of function of variability density distribution and the value of argument at maximum point – with moda (Mo). Then in function of variability density distribution in 2% level from Amo founded minimal (xmin) and maximal (xmax) function values.
Variational sweep (MxMDn) – is difference between minimal and maximal values of R-R intervals dynamical row:
(10)
the relation of minimal and maximal values of R-R intervals
(11)
53
∆ x 0.025 5.83y+
h1N
50∆ x
NClass
⋅
MxRMnxmax
xmin
MxDMn xmax xmin−
stress-index (index of regulatory systems tension - SI) counted by division moda amplitude to double multiplicative of moda and sweep
(12)
Autocorrelaive analisys Correlation coefficient after 1st shift (CC1):CC1=r0,1, where r0,1 is correlation coefficient which counted by the calculation
of autocorellatonal function with shift value – 1 second. Autocorrelaive function is built on values of correlation coefficient row betweeen source dynamical row xi and new rows, which we got by subsequent shifting by one mesure.Correlative coffi-cient is counted by formula:
, k=0,…,m-1(13)where k is shift step number, m is shift steps quantity (m=128 when step
∆t=250 ms).number of shifts to first zero level of correlative cooefficients (CC0):
(14)
Spectral analysisFor spectral analysis of dynamic rows of cardiointervals proposed using of
non-parametrical methods, based on using of straight Fourie transformation of func-tion x(f) in frequent distribution (specter). In realisation of this algorithm on PC usu-ally use discrete Fourie transformation and particularly fast Fourie transformation, using next 2 formulas:
54
r0 k,
m
1
m
i
xi xi k+⋅( )∑=
⋅
1
m
i
xi
1
m
i
xi k+∑=
⋅∑=
−
m
1
m
i
xi( ) 2∑=
⋅
1
m
i
xi∑=
2
−
m
k 1+
m k+
i
xi( ) 2∑=
⋅
k 1+
m k+
i
xi∑=
2
−
⋅
CC0 k∆ t
1000⋅ r0 k, 0,
SIAmo 100%⋅
2 Mo s( )⋅ MxDMn s( )⋅
(15)
(16)
Here
N is a number of counting out, Δt is time interval between the ones, Δω is spectrum stem in the frequency domain, that calculates by the formula:
(17)
, and T – analyzing signal time interval, a so-called record length or main order:
(18)
Spectrum (15) is dissymmetrical (two-sided) relative to its central point l=(N-1)/2, that is: Xl=XN-1, that’s why for its graphic representation and following investigation it’s enough first (N-1)/2 amplitudes (one-sided) spectrum. When switching to the one-sided spectrum from the two-sided one its amplitudes should
55
Xl
0
N 1−
k
xk e jl∆ ω k∆ t−⋅ ∆ t∑=
l 0, 1, .., N 1−,
xk1
2π0
N 1−
l
Xl e jl∆ ω k∆ t−⋅ ∆ ω∑=
k 0, 1, .., N 1−,
xk x k∆ t( )
Xl X l∆ ω( )
∆ ω2πT
T N 1−( ) ∆ t⋅
be normalized by multiplication on 2 (power spectrum is normalized by multiplica-tion on 2)
The upper bound of analyzing spectral band is determined by the numeral-ization frequency of signal fs=1/Δt and equal fs/2, and the lower bound is equal fre-quency resolution1/T. Value 1/T is also called the main circular frequency. Spec-trum analysis results frequency range from 1/T to f/2 is called spectral band width.
For getting well smoothed (interpolated) spectrum by brief signal realiza-tion and for the spectral peak frequency evaluation accuracy increasing the basic temporary consequence supplement with nulls is executed. It results in appearance of m=n/N intermediate values in the spectrum, where n – number of nulls added; N – basic signal values number in temporary realization. However it is possible to in-crease frequency resolution only due to analyzing signal section duration increasing, but certainly not due to supplement with nulls.
In general case for the (15) execution it is necessary to calculate N2 product of xkFN
m, where FNm=(e-jlΔωkΔt)m – multiplication factor (m=kl).
SPF is calculated by a number of discrete values xi, I=1,2,…,N, with ob-tained function x(t) quantification method by the next algorithm:
1. The five-minutes record division into three segments;2. Function x(t) centering in each segment with regard to the average value
(constant component elimination) and simultaneous weighing of it (fon Hann window application) according to formula: ;
(19) , where xi, xi are basic and centralized-weighed signals amplitudes,x is average value, calculated according to formula (2), and W – fon
Hann window, that in time domain is described like squared cosine function:
(20)3. Value series xi, I=1,2,…,N supplement in every segment with nulls to the
nearest number “two in power”. In compliance with agreements (ch. 2) a three-minute segment consists of 720 readings, what should be supplied with nulls up to 1024 readings.
4. Fourier conversion of the value series, i=1,2,..,N in every segment according to formula (15) using fast Fourier transform.
5. Spectrum Xi amplitudes normalization by multiplication by √2.6. SPF computation according to formula:
(ms2) . (21)
56
W i 0.5 1 cos 2πi
N 1−2
−
N 1−
⋅
+
⋅ i 0, 1, 2, .., N 1−,
PlN2
Xl( ) 2⋅l 0 1, ..,
N 1−( )2
,
7. SPF segments linear averaging-out
Value Pi(ω) is also called power falling on frequency Δω unit on the fre-quency ω. Total power is equal sum of powers falling on all units of frequencies Δω.
On a diagram power is showed in frequency k/T values that changes from 1/T to 1/(2 Δt). In Russia the agreement of inverse scale on horizontal axis use with-out periodogram modification. At that abscissas are measured in run length.
Spectral analysis indexes calculation is kept in four frequency ranges ΔfHF, ΔfLF, ΔfVLF, and ΔfULF;
high-frequency rippling HF in range:0.4-1.5 hertz (2-6.6 sec)low-frequency rippling LF in range:0.15-0.04 hertz (7-25 sec)very low frequency rippling VLF in range:0.04-0.015 hertz (25-66 sec)ultralow-frequency rippling ULF in range:0.015-0.003 hertz (66-333 sec) Next values are to be calculated by the spectrum evaluations: HF, LF,
VLF, ULF – spectrum powers in frequency ranges ΔfHF, ΔfLF, ΔfVLF, and ΔfULF cor-respondingly.
In every of frequency ranges ΔfHF, ΔfLF, ΔfVLF, and ΔfULF harmonics powers spectrum evaluations maximal values (HFmx, LFmx, VLFmx, and ULFmx) are to be found.
Spectrum HF power (total power in the frequency range ΔfHF) is calculated according to formula:
,(ms2) (22)
, where LHFr and LHFl are numbers of spectrum evaluations corresponding to range ΔfHF scope.
Powers of spectra LF, VLF, ULF (in frequency ranges ΔfLF, ΔfVLF, and ΔfULF) are calculated in much the same way.
Total spectrum power:TP=HF+LF+VLF+ULF;
HFt,LFt, VLFt, ULFt – spectrum peaks maximal (dominating) periods values in corresponding frequency ranges;
Average spectrum power in frequency range ΔfHF:
57
HF
LHFl
LHFr
j
P j∑=
HFavHF
LHFr LHFl−( )
(25)
Average spectrum power in frequency range ΔfLF:
(26)
Average spectrum power in frequency range ΔfVLF:
(27)
Average spectrum power in frequency range ΔfULF:
(28)
Spectrum power in frequency range ΔfHF percentage wise to all range:
(29)
Spectrum power in frequency range ΔfLF percentage wise to all range:
(30)
Spectrum power in frequency range ΔfVLF percentage wise to all range:
(31)
Spectrum power in frequency range ΔfULF percentage wise to all range:
(32)
Centralization index
(33)
58
LFavLF
LLFr LLFl−( )
VLFavVLF
LVLFr LVLFl−( )
ULFavULF
LULFr LULFl−( )
HF%HFTP
100%⋅
LF%LFTP
100%⋅
VLF%VLFTP
100%⋅
ULF%ULFTP
100%⋅
59
60
61
62
63
64
65
66
67